Location:Home > Engineering science > Computer Science > Computer technology > The Research of the Control Algorithm for Nonlinear System Based on RBF Neural Network

The Research of the Control Algorithm for Nonlinear System Based on RBF Neural Network

Downloads: []
Tutor: XuZhiHong
School: Hebei University of Technology
Course: Computer technology
Keywords: uncertain nonlinear system Lyapunov stability theory,adaptive control,robustH_∞c
CLC: TP183
Type: Master's thesis
Year:  2013
Facebook Google+ Email Gmail Evernote LinkedIn Twitter Addthis

not access Image Error Other errors

Almost systems in the practice have the characteristics of nonlinear and uncertainties,mostly, the uncertainties come from the error of system modeling, varying parameters andexternal disturbances. Based on the above reasons, there have the big difficulties fornonlinear system control. Robust control and adaptive control are main approaches whichcan cope with the uncertainties, and have been made a great deal of development. In thispaper, to handle the nonlinear systems which have construct, parameter uncertainties andexternal disturbances, we use the robust control technique, adaptive control technique,backstepping control technique and neural network technique as main research tools; toexplore the problem how to design the robust adaptive control algorithm for uncertainnonlinear system. The primary work of this paper is given as follows:To solve the uncertainty control problem of nonlinear system, this paper designs a setof designed algorithm to the two kinds of uncertain nonlinear systems according to theLyapunov stability theory, based on the robust backstepping control algorithm, the robustadaptive backstepping control algorithm by the robust control, the adaptive control andbackstepping control technique. The proposed algorithm can guarantee the boundedness,stability and the performance of control output of the closed-loop system. At last, this paperused the traditional uncertain nonlinear system-two link manipulator as the simulationobject, the given simulation results show that the proposed algorithm is effective.To aim at the nonlinear systems which exists the unknown external disturbances andinner uncertainties, we present the robustH_∞control algorithm which is based on RBFneural network disturbance observer to reduce the requirement of disturbances for systems.The external disturbances, inner uncertainties and the cross coupling of the systems arecombined as multiple disturbance, based on RBF neural network, we design the disturbanceobserver to approximate the multiple disturbance. The proposed control algorithmcombines the neural network control; robust control andH_∞control, which guarantee thestability and ultimately uniformly boundedness of the closed-loop system. The simulationresults show the proposed control algorithm can track the reference trajectory well; theaffection of multiple disturbances is suppressed. To resolve the compute complexity of thedisturbance observer, we use the strong approximation ability of neural network tocompensate the affection of control performance from the uncertainties, based on RBFneural network. we design the nonlinear system adaptive output feedback control algorithmwhich needn’t add the state observer in the system control. Based on analysis of stabilitytheory, the proposed control algorithm can confirm the stability of system and theboundedness of output signal. The simulation results indicate that the proposed algorithmcan guarantee the system to obtain the desired control performance.
Related Dissertations
Last updated
Sponsored Links
Home |About Us| Contact Us| Feedback| Privacy | copyright | Back to top